OK, in response to various concerns, I decided to tighten up me procedures. I used a different section of the same negs, closer to infinity - or at least down to the end of the block. I was also more careful with the levels, and didn't apply any sharpening in Photoshop. These samples are presented at 200% actual size, resampled using bicubic interpolation in Photoshop. The jpegs are at 100%.

On my screen (19", 1280x1024) these would be sections of a 72" x 104" image - that's 6 by 9 feet, or for our continental friends, 183 x 264 cm.

In this series, there isn't any significant difference at all. In fact, if I squint the Hexar image looks a hair sharper. As far as I'm concerned, this case is closed. Nailed, glued and screwed shut.

Hexar

Leica

-- Paul Chefurka (paul@chefurka.com), May 19, 2002

Answers

Well I guess the Hexar RF problem was cured in later production. Just wonder what the approximate "cut-off" point is, both chronological and in serial number.

Paul, when did you get yours and what serial number does it have?

-- Andrew (mazurka@rocketmail.com), May 19, 2002.

Wow what a lens wow ! Thank for your test, Mr Chefurka. Do one at minmum focus distance.

-- Richard Brown (rubyvalentine@earthlink.net), May 19, 2002.

Andrew - I got mine in late 2000. the serial number is 1448755.

Richard - I think I'm all tested out for the moment. This has been an iteresting one, though. You're right about the lens. To get that kind of resolution out of a 28/2.0 wide open is frankly astonishing even to this jaded old fart.

-- Paul Chefurka (paul@chefurka.com), May 19, 2002.

Hi Paul:

Good test, but I wouldn't be so hasty as to glue, screw and nail the case shut just yet... In your earlier post you stated that infinity is a better test of this issue -- I don't think so. As we approach infinity, our DOF lengthens considerably and hence a back-focus problem will mitigate itself to a certain -- perhaps imperceptible -- degree. Where it will crop up continuaously is in the closer subject, wide aperture shot. Hence, I suggest to you that a test subject -- and a 3 dimensional subject will illustrate it best -- shot at say 6' distance at f2 (like we might do with our Leica's in "street" mode) will offer the most telling example.

Should you decide to perform this test, if you focus carefully on the subject's eyes, I think you will find that:

1) With the Leica, the sharpest focus point will be very near their eyes, and

2) With the Hexar, the best focus point will be somewhat closer, like their nose.

Now, the eyes may still be acceptably sharp with the Hexar, but not as critically sharp as the nose... THIS is the issue I had with the Hexar and back focus. And this problem will drop to insignificance as the subject distance increases, and the "dimensionality" of the subject decreases (flattens); such as in landscape or scenic type shots.

Cheers,

-- Jack Flesher (jbflesher@msn.com), May 19, 2002.

Jack, the problem with a test like you propose is that you're at the mercy of the mechanics. The cam follower arm is a finicky piece that can easily be misaligned, and the effects of any malajustment typically become greater at closer focus. While a closeup test will show differences between the systems, it will not prove that these are due to back focus problems - you're at the mercy of both the mechanics and yo're own optical acuity. That's why I wanted to run an infinity test in the first place - to take the focussing mechanism out of the equation.

The cam follower can be adjusted to provide proper tracking between close focus and near infinity, but at infinity you are at the mercy of the system geometry. If the system focusses properly at infinity, it can probably be adjusted to focus correctly at closer distances. If the infinity focus is off, there is nothing you can do short of modifying each lens' helical to compensate for the incorrect body depth.

I still think this test proves what it set out to do - this Hexar and this Leica both focus properly at infinity with this lens. There doesn't seem to be any significant FFD difference between these bodies. If I now discovered a close focussing error in either camera I'd expect it could be adjusted out by mods to the camera's focussing mechanism.

In a nutshell - IMO close focussing tests say more about the state of the mechanics and the tester's eyes than about the camera's design.

-- Paul Chefurka (paul@chefurka.com), May 19, 2002.

Paul:

I understand what you are saying, and I understand your point, and it may be valid for you and the type of shots you normally take...

But, if you adjust out the close focussing issues on the Hexar for example, its infinity focus will now be off, and the ONLY way you can properly correct this problem is to have the lens cam reground to match Hexar specs...

Plus, if you do the test as I have laid it out, and repeat it a few times, and get the same or similar results each time, you can then be pretty sure it is a compatibility issue, no matter how minor for your particular uses. And long back-focus issues will generate a true focus point closer than the indicated focus point in the final image. Always.

Cheers,

-- Jack Flesher (jbflesher@msn.com), May 19, 2002.

no diferences now, may be mask by DOF, previous samples at shorter distance had a diference, is quite logical, you get to a point where DOF mask milimetric diferences, for mr first sample was enough to get your idea Paul, thank´s for you sunday work man.

-- r watson (al1231234@hotmail.com), May 19, 2002.

paul -- assuming that the hexar positions the lens .4 mm further away from the film plane than it should, this translates into a slight focus adjustment back from infinity -- to somewhere between 30m and inf (johann was way off in his calcs). since the house in the foto is not at infinity, but closer, this difference mite account for the sharper hexar photo. i hate to say it, but the only real test is either to shoot something near infinity (a distant skyline is a good test), or to shoot something very close. the true infinity shoot is better for all the reasons you mention above. there is a big difference in the sharpness of the above fotos -- and the result is exactly what you'd expect if the hexar back focus was slitely long. p.s. the original pair favored the leica -- do you now put this down to contrast/digital sharpening eccentricities?? YOU'RE A TROOPER!!

-- roger michel (michel@tcn.org), May 19, 2002.

When I get the chance here's what I'm going to do: attach a Visoflex to an M6 and Hexar, with a 90 Elmarit-M lens mounted to it, and place them on my copy stand and photograph a postage stamp at f/2.8. There is almost no DOF and since the focus will be determined on the Viso's ground glass which is calibrated to the Leica register, if the Hexar register is not the same the image should be soft. Sound logical?

-- Jay (infinitydt@aol.com), May 19, 2002.

Roger,

For all practical purposes I consider the resolution of those sections to be identical. If the FFD were off as much as anyone has hypothesized, I'd expect to see evidence of it in these images, and frankly, I don't. I can see the mullions in the windows of a house that's over 600 feet away, at f/2.0 with a 28mm lens, on both bodies. That's close enough for government work, and good enough for me.

I'm going to stop fretting about it and take the Hexar out for some real photography.

-- Paul Chefurka (paul@chefurka.com), May 19, 2002.

I have a roll of film somewhere where I borrowed a friend's leica 50 and took half a roll of pictures of a tape measure at a slight angle from the camera. I focussed the camera on a known spot on the tape measure and shot a bunch of frames from a tripod with the Leica 50, a Leica 35/2 and the Konica 50.

Under a 15x loupe, the expected parts of the photos were all sharp, even at closest focus wide open with every lens.

This was with an early 2001 Hexar, and it seemed pretty conclusive at the time.

-- Pete Su (psu@kvdpsu.org), May 19, 2002.

Oh yeah. I also have tons of pictures taken at near infinity with my Hexar and a Leica 35/2. And they are all sharp too.

This whole issue just seems like quibbling.

-- Pete Su (psu@kvdpsu.org), May 19, 2002.

sounds like a wise decision!! government work?? you're not one of those guys in the black heliocopters are you??

-- roger michel (michel@tcn.org), May 19, 2002.

Oh darn. Now we're going to have to make you disappear. I really wish you hadn't figured that out...

-- Paul Chefurka (paul@chefurka.com), May 19, 2002.

Jay:

You have offered the most logical testing proceedure yet. Please share your results when you get them in.

Cheers,

-- Jack Flesher (jbflesher@msn.com), May 19, 2002.

The only test that will nail his argument shut is an accurate measurment of the camera body - film tests are open to variations in too many factors to be conclusive ( film flatness variations between exposure, exposure etc). All Paul's test shows is that one lens is 'apparently' slightly sharper on one body. M bodies can be wrong too - yes it's true!

-- Johann Fuller (johannfuller@hotmail.com), May 20, 2002.

Substitute This for His in the above statement -sorry its early in the morning and my keyboard dexterity is not up to speed yet!

-- Johann Fuller (johannfuller@hotmail.com), May 20, 2002.

Dear Paul,

You are a trooper, good friend. Thank you for this last test. I'll sleep well now (it's 1 a.m. here in Japan). I think the case is now nailed shut and buried. R.I.P.

If in fact there is a problem it is a small one. If it was not a small one but a big one there would be no arguments.

I'm publishing some photos made with my Hexar RF and Leica lenses if that will make anyone feel better.

Good night,

Alex

-- Alex Shishin (shishin@pp.iij4-u.or.jp), May 21, 2002.